| The safe and efficient excavation of coal mine roadway plays a very important role in coal and rock mining.However,the roadway environment is complex and the safety hazards are large.The traditional roadway deformation detection method is inefficient,time-consuming,and requires manual in-depth field measurement.Therefore,the use of robots to participate in the inspection of coal mine roadways is the future development direction.With the development of 3D laser scanning technology,lidar can quickly and accurately obtain environmental information and is widely used in various industries.Therefore,this thesis takes the mobile inspection robot as the platform,and realizes the autonomous walking and roadway deformation detection of the inspection robot in the roadway environment by carrying 3D laser radar.The details are as follows :(1)The inspection robot platform and system model are built.Firstly,based on the requirements analysis of intelligent detection in coal mine roadway,this thesis designed an experimental platform for a tracked mobile inspection robot and determined the hardware and software system scheme for the robot.Then,according to the overall scheme of the system,this thesis established a targeted system model for the inspection robot,which provided a foundation for the robot’s motion control and roadway environment scanning.(2)A hybrid path planning method based on improved A * algorithm and DWA algorithm is proposed and verified by experiments.Firstly,based on the global path planning A* algorithm,we propose an improved heuristic function that includes target weight information to address problems such as excessive path turning points and nonsmooth paths in traditional A* algorithm planning.We also use Floyd algorithm and Bspline curve to eliminate redundant nodes and smooth paths.Secondly,we integrate the global path planning A* algorithm with the local path planning DWA algorithm,extract key nodes information from the improved A* algorithm global path,and set the extracted global path nodes as local target points for DWA algorithm,thereby dynamically planning the path to the target points.Experimental results show that the proposed hybrid path planning method combining the improved A* algorithm and DWA algorithm has higher efficiency and better safety,and can meet the motion requirements of coal mine inspection robots in roadway.(3)The key technology of roadway deformation detection based on laser scanning is studied.Firstly,we explained the working principle of point cloud denoising and downsampling methods,and proposed a point cloud denoising method that combines statistical filtering and median filtering,as well as a voxel grid downsampling method based on density constraint and centroid near points to address the shortcomings of existing denoising and downsampling methods.Then,we used the bi-directional projection method to extract the centerline of the roadway,segmented the point cloud data using Cloud Compare point cloud processing software,extracted the cross-section of roadway,and fitted the contour curve.Finally,we introduced a method to extract the deformation of roadway based on the processed point cloud data,and perform deformation detection and analysis.(4)In order to verify the feasibility of the overall scheme,the experimental platform of inspection robot is used to carry out the experimental study of autonomous walking and the application analysis of roadway deformation detection.Firstly,we verified the performance and practical application effect of the proposed path planning algorithm using the inspection robot platform.Then,we processed the simulated roadway point cloud data obtained by laser scanning,extracted the deformation amount,and performed deformation analysis of roadway.This thesis contains 58 figures,19 tables and 115 references. |
PDF Full Text Request